This cross-sectional study was carried out in the administrative sectors of the Suez Canal University campus in Ismailia, Egypt, from July 2022 to February 2023. Women employees and workers aged ≥ 45 years working in the administration offices of the university and faculties who had fulfilled the criteria of the target population of perimenopausal, and postmenopausal women were recruited. The perimenopausal group included women who had experienced persistent cycle irregularity, with a difference in cycle length of at least 7 days on two or more occasions within the previous 10 cycles, and/or amenorrhea for at least 60 days but no more than 12 months within the last year. The postmenopausal group included women who had experienced their final menstrual period more than one year prior [19]. Exclusion criteria included the following: having undergone the removal of one or both ovaries or a hysterectomy; having polycystic ovary syndrome or premature ovarian failure; experiencing premature menopause (defined as cessation of menstrual periods before the age of 40); using exogenous hormones; experiencing amenorrhea following cancer treatment; confirmed pregnancy; and a history of CVD, defined as being diagnosed with or treated for CVD.

The sample size was calculated using the G*Power program for Windows [20]. Using linear multiple regression (fixed model R2 deviation from zero), a total calculated sample of 366 was estimated to be enough for testing the strength of association between potential predicting variables and underestimation of perceived CVD risk among perimenopausal and postmenopausal women. This calculation was based on a multiple linear regression model involving at least 20 predictors, at effect size f2 = 0.06, α error = 0.05, and power = 0.80. This number was increased to 421 to allow for an expected non-response rate of 15%. The sample size of participants from each administrative sector was determined based on the proportionate size of the eligible female population in each sector. Using a random integer generator from a reputable software program (RANDOM.ORG's integer generator), women were randomly selected from a list of coded-eligible participants and invited to participate in the study. The study’s aims and procedures were explained to the women. Those who accepted to participate signed consent forms, resulting in a final total of 390 participants and a response rate of 92.6% (Fig. 1).

Flow diagram of study participant enrollment

Participants’ self-reported data were collected using a structured interview questionnaire composed of two parts: the first part comprises the socio-demographics data including age, education, marital status, monthly income, and occupation and medical history data of study participants. Smoking status, experiencing chest pain, and shortening of breath were assessed using one question for each with a dichotomous response (Yes or No). A “smoker” is referred to as someone who smokes daily or occasionally, and a non-smoker is referred to as someone who has stopped smoking or has never smoked. Experiencing chest pain refers to feeling tingling or pain in the chest, while shortness of breath refers to difficulty breathing, particularly when going upstairs or walking quickly. Having diabetes, hypertension, or hyperlipidemia was defined as being diagnosed and/or treated.

The second part comprises the participants’ self-perceived risk of developing CVD (myocardial infarction and stroke) in the next 10 years, their self-perception regarding their general health, and their awareness of factors that can increase the risk of developing CVD. The self-perceived risk of developing CVD in the next 10 years was assessed using a categorical measure with a 3-point Likert scale: low, moderate, and high degree of risk, using the question “How do you perceive your risk of having a myocardial infarction or stroke in the next ten years?’’. Participants were also asked to rate their current health status on a scale from 0 to 10 with 5 equal intervals, rated as very poor, poor, fair, good, and very good [18]. Awareness of factors that can increase the risk of experiencing CVD was assessed using ten items. These items included obesity, diabetes, hypertension, hyperlipidemia, aging/menopause, hereditary factors and behavior related factors as unhealthy diet, lack of exercise, smoking, and drinking alcohol which were considered to increase the risk of developing CVD [4, 21]. Responses were dichotomized into “Yes” (1) or “No” (0), and a total awareness score (ranging from 0-10) was obtained by summing across all items, with higher scores indicating a higher level of awareness.

Perceived stress was assessed based on its duration and intensity. To measure the perceived duration of stress, participants responded to a question which asked: "Which of these statements best represents your exposure to psychological stress during the past five years?" The scale included options ranging from short periods of stress ("never experienced stress," "experienced some periods of stress during the past five years") to moderate periods ("experienced long periods of stress that lasted one year during the past five years"), and long periods ("constant stress during the last year," "constant stress during the past five years") [18].

To evaluate the perceived intensity of stress over the previous month, the validated Arabic version of the Perceived Stress Scale (PSS-10) was used [22]. This scale consists of 10 questions with responses on a five-point Likert scale, ranging from "never" (0) to "very often" (4). The total score on the Arabic PSS-10 is calculated by summing the scores of all items. Stress levels were categorized as low (0–13), moderate (14–26), and severe (27–40), with higher scores indicating greater perceived stress. The Arabic PSS-10 has demonstrated concurrent validity and acceptable reliability [22].

Health literacy, defined as the ability to find, understand, and use health information to make informed decisions and take care of one's own health [23], was assessed using the validated Arabic 16-item short version of the European Consortium for Health Literacy Questionnaire (HLQ-EU-16) [24]. The questionnaire items address self-reported difficulties in tasks concerning decision-making in health care, disease prevention, and health promotion using a 4-point Likert scale ranging from 1 (very difficult) to 4 (very easy). Total HL scores (with a range from 0 to 16) were obtained by dichotomizing answers to 0 (very difficult/fairly difficult) and 1 (fairly easy/very easy). Scoring establishes three HL categories: 0–8 = inadequate, 9–12 = problematic, and 13–16 = sufficient HL. For the purpose of the analysis, the HL variable was dichotomized into "inadequate/problematic" and "sufficient" based on the findings of a previous study [18] and to align with determining the influence of sufficient HL compared to lower HL levels on the underestimation of CVD risk. The questionnaire had high internal consistency, and construct validity [24]. Numeracy, defined as the ability to understand and use numerical information for health decision-making [25], was assessed using the Short 3-item Version of Subjective Numeracy Scale (SNS-3), with a five-point Likert scale (with higher scores indicating high numeracy), from which a mean summary score was calculated. The SNS-3 demonstrated both high internal consistency and good evidence of criterion and construct validity [26].

Participants’ clinical data included blood pressure and body mass index (BMI) measures. The blood pressure was measured using the standard mercury sphygmomanometer in the left arm. The average of two measures (≥ 2 min apart) was obtained after the participant had been sitting for ≥ 5 min. Participants’ BMI was calculated using the formula [BMI = weight (kg) / height (m2)], with height and weight measured using a height measuring stand and a weighing scale, respectively. Measurements were taken with participants wearing clothing without shoes.

The 2019 World Health Organization (WHO)/CVD risk-non-laboratory-based prediction chart, tailored for the North Africa and Middle East Region, was used to estimate participants' 10-year risk of experiencing a major cardiovascular event, such as a heart attack or stroke. This chart, which follows WHO guidelines, assesses risk based on five factors: age, sex, smoking status, blood pressure, and BMI. Risk is categorized as low (less than 10%), moderate (10% to less than 20%), or high (20% or more) [27]. Participants self-perceived risk accuracy was determined by comparing their self-assessed risk of developing CVD in the next 10 years with the risk predicted by the WHO chart.

Reliability analysis of the current study sample revealed Cronbach’s alpha of 0.947 for the PSS-10, 0.963 for the HLQ-EU-16 questionnaire, and 0.708 for the SNS-3. The whole questionnaire was pre-tested on 20 women who were not included in the study. The pilot testing did not necessitate any changes to the questionnaire construction or wording. The average time required to complete the questionnaire ranged from 10 to 15 min.

Data were assessed for normality using the Kolmogorov-Smirnov test, which indicated that continuous variables were not normally distributed. Consequently, medians and interquartile ranges (IQR) were used for descriptive statistics. Categorical variables were summarized using frequencies and percentages. The Kappa test was used to detect the degree of agreement between self-perceived and predicted CVD risk. A two-step statistical approach was employed to identify factors that predict underestimation of CVD risk. Firstly, significant variables were identified through simple logistic regression analysis (p < 0.05). Secondly, backward stepwise multiple logistic regression with likelihood ratios (LR) was used to create a model that included only the most relevant and statistically significant predictors. All statistical analyses were performed using SPSS 25 (Armonk, NY: IBM Corp.).